https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
Mark Thomas <[EMAIL PROTECTED]> changed:
What|Removed |Added
CC||[EMAIL PROTECTED]
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #58 from Mark Thomas <[EMAIL PROTECTED]> 2008-08-15 00:23:15 PST
---
Just for the record, the 5.5.x fixes to the connector module were sufficient to
fix this issue for 4.1.x as well.
--
Configure bugmail: https://issues.
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
Mark Thomas <[EMAIL PROTECTED]> changed:
What|Removed |Added
Status|ASSIGNED|RESOLVED
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #56 from Mark Thomas <[EMAIL PROTECTED]> 2008-07-09 14:32:46 PST
---
*** Bug 45349 has been marked as a duplicate of this bug. ***
--
Configure bugmail: https://issues.apache.org/bugzilla/userprefs.cgi?tab=email
---
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
Mark Thomas <[EMAIL PROTECTED]> changed:
What|Removed |Added
CC||[EMAIL PROTECTED]
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
Paul Ryan <[EMAIL PROTECTED]> changed:
What|Removed |Added
CC||[EMAIL PROTECTED]
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
Andrew Savory <[EMAIL PROTECTED]> changed:
What|Removed |Added
CC||[EMAIL PROTECTE
Mark Thomas wrote:
[EMAIL PROTECTED] wrote:
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #52 from Rainer Jung <[EMAIL PROTECTED]> 2008-04-28
11:53:54 PST ---
As soon as svn will be back to read/write, I'll add the patches to the
TC 5.5
STATUS file.
You should find
[EMAIL PROTECTED] wrote:
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #52 from Rainer Jung <[EMAIL PROTECTED]> 2008-04-28 11:53:54 PST
---
As soon as svn will be back to read/write, I'll add the patches to the TC 5.5
STATUS file.
You should find some entries already t
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #52 from Rainer Jung <[EMAIL PROTECTED]> 2008-04-28 11:53:54 PST
---
I ported the patches for 44494 applied until now to TC 6.0 back to TC 5.5.
I reviewed all differences for the classes in o.a.tomcat.util.buf and
o.a.c.co
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
Rainer Jung <[EMAIL PROTECTED]> changed:
What|Removed |Added
Attachment #21872|Patch for TC 5.5 Part 1 - |Patch for TC 5.5
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #51 from Rainer Jung <[EMAIL PROTECTED]> 2008-04-28 11:48:39 PST
---
Created an attachment (id=21873)
--> (https://issues.apache.org/bugzilla/attachment.cgi?id=21873)
Patch for TC 5.5 Part 2 - container (InputBuffer)
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #50 from Rainer Jung <[EMAIL PROTECTED]> 2008-04-28 11:47:29 PST
---
Created an attachment (id=21872)
--> (https://issues.apache.org/bugzilla/attachment.cgi?id=21872)
Patch for TC 5.5 Part 1 - connectors (InputBuffer)
Bac
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #49 from Suzuki Yuichiro <[EMAIL PROTECTED]> 2008-04-20 17:43:44
PST ---
Oops, I forgot to reset InputBuffer.readAheadLimit in the recycle method
of the patch, though the result may be O.K. with or without it.
If you use th
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
Suzuki Yuichiro <[EMAIL PROTECTED]> changed:
What|Removed |Added
Attachment #21683|0 |1
is
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #47 from Suzuki Yuichiro <[EMAIL PROTECTED]> 2008-04-18 21:46:52
PST ---
I tried current 6.0.x. It looks working fine at all test cases.
Thank you very much for the fix.
But there is a few comments.
1. readAheadLimit value
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #46 from Mark Thomas <[EMAIL PROTECTED]> 2008-04-11 14:06:49 PST
---
OK. That light up ahead might actually be the end of the tunnel rather than yet
another train.
I believe that all necessary patches to fix all issues ide
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #45 from Suzuki Yuichiro <[EMAIL PROTECTED]> 2008-04-10 21:26:03
PST ---
Hi,
The behavior of the test case with default values is as follows.
1. It posts 8192 multibytes characters.
2. The posted data is read up to 4096
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #44 from Rainer Jung <[EMAIL PROTECTED]> 2008-04-10 15:29:58 PST
---
Next mid-air, BZ is obviously not IM :)
When I use the test war, the garbage test still seems to fail, even with
default maxHttpHeaderSize. The behaviour
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #43 from Filip Hanik <[EMAIL PROTECTED]> 2008-04-10 15:21:07 PST
---
hi Rainer, both patches do the exact same thing, one explicit one implicit, I'd
vote +1 for either or
Filip
--
Configure bugmail: https://issues.apach
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #42 from Filip Hanik <[EMAIL PROTECTED]> 2008-04-10 15:20:00 PST
---
If we don't want the simple fix of growing the CharChunk, (since CharChunk will
always be <=maxHttpHeaderSize, a character is at least one byte)
then the
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #41 from Rainer Jung <[EMAIL PROTECTED]> 2008-04-10 15:15:06 PST
---
Hi Filip,
I didn't yet read your comment, because we both worked in parallel (mid-air
collission). I'm posting my finding nevertheless, so we can look fo
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #40 from Filip Hanik <[EMAIL PROTECTED]> 2008-04-10 14:56:29 PST
---
I think I might have found the problem,
I've been unable to reproduce the error using the NIO/APR (APR below) connector
with setting
I reproduce t
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #39 from Rainer Jung <[EMAIL PROTECTED]> 2008-04-10 11:41:32 PST
---
And the stack in realWriteChars is:
at
org.apache.catalina.connector.InputBuffer.realWriteChars(InputBuffer.
java:335)
at org.apache.tomc
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #38 from Rainer Jung <[EMAIL PROTECTED]> 2008-04-10 11:25:43 PST
---
Since most people are deeper in the code than I, here's what i get if I
instrument o.a.c.connector.InputBuffer with a couple of log statements and post
th
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #37 from Rainer Jung <[EMAIL PROTECTED]> 2008-04-10 08:24:21 PST
---
Using the war and maxHttpHeaderSize="16384" I did the following test case:
Characters size: 17000___
Use MultiByte Character: ( ) yes (
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #36 from Rainer Jung <[EMAIL PROTECTED]> 2008-04-10 08:15:03 PST
---
I'll try that.
For the sake of completeness, here's the exception stack:
java.io.IOException
at
org.apache.catalina.connector.InputBuffer.reset(
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #35 from Remy Maucherat <[EMAIL PROTECTED]> 2008-04-10 08:05:13
PST ---
You've got to be able to reproduce this using the test webapp provided in this
report (the line length is configurable, and there's a readLine test). I
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
Ruediger Pluem <[EMAIL PROTECTED]> changed:
What|Removed |Added
CC||[EMAIL PROTECT
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #34 from Rainer Jung <[EMAIL PROTECTED]> 2008-04-10 01:59:00 PST
---
I tested recent trunk and also 6.x with the latest STATUS file patches for this
issue applied.
When we change maxHttpHeaderSize from the default to somet
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #33 from Mark Thomas <[EMAIL PROTECTED]> 2008-04-07 15:50:41 PST
---
I have a fix for this now. Thsi fix and Remy's fix above have been proposed for
6.0.x and 5.5.x
--
Configure bugmail: https://issues.apache.org/bugzill
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #32 from Mark Thomas <[EMAIL PROTECTED]> 2008-04-07 14:27:43 PST
---
The mark/reset test case still fails. I'm looking into it but it is taking me
time to figure out exactly where the error is.
--
Configure bugmail: http
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #31 from Remy Maucherat <[EMAIL PROTECTED]> 2008-04-07 08:10:29
PST ---
I found a problem with the buffer resizing in conjunction with mark/reset.
Index: java/org/apache/tomcat/util/buf/CharChunk.java
=
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #30 from Mark Thomas <[EMAIL PROTECTED]> 2008-04-05 10:49:43 PST
---
The multi-byte read patch has been applied to 6.0.x and will be in the next
release.
--
Configure bugmail: https://issues.apache.org/bugzilla/userprefs
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #29 from Mark Thomas <[EMAIL PROTECTED]> 2008-03-30 14:19:03 PST
---
I have tested Remy's patch (less the one bad line) and it does fix the various
multi-byte read test cases.
The patch has been applied to trunk and propos
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #28 from Remy Maucherat <[EMAIL PROTECTED]> 2008-03-26 05:44:40
PST ---
(In reply to comment #26)
> > cb.setLimit(cb.getStart() + cnt);
> > cb.append( result, 0, cnt );
>
> cb.append() doesn't modify cb.start
> and cb.star
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
Suzuki Yuichiro <[EMAIL PROTECTED]> changed:
What|Removed |Added
Attachment #21705|0 |1
is
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #26 from Suzuki Yuichiro <[EMAIL PROTECTED]> 2008-03-25 21:11:57
PST ---
(In reply to comment #23)
> cb.setLimit(cb.getStart() + cnt);
> cb.append( result, 0, cnt );
cb.append() doesn't modify cb.start
and cb.start is set
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #25 from Remy Maucherat <[EMAIL PROTECTED]> 2008-03-25 19:34:24
PST ---
The first step is to reproduce the problem (which is easy for me with TC 6.0
trunk and the latest version of the test case). After fixing the update of
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #24 from Filip Hanik <[EMAIL PROTECTED]> 2008-03-25 19:04:21 PST
---
when I ran the test:
it fails consistently with tomcat/6.0.x/trunk
after updating tomcat/trunk to the latest(same fix as in 6.0.x)it also fails
consiste
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #23 from Remy Maucherat <[EMAIL PROTECTED]> 2008-03-25 18:27:12
PST ---
Ok, first of all thanks a lot for the test case (I wouldn't be able to write a
good multibyte test to save my life).
As with the original bug, the pro
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
Suzuki Yuichiro <[EMAIL PROTECTED]> changed:
What|Removed |Added
Attachment #21700|0 |1
is
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
Suzuki Yuichiro <[EMAIL PROTECTED]> changed:
What|Removed |Added
Status|NEEDINFO|ASSIGNED
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #20 from Suzuki Yuichiro <[EMAIL PROTECTED]> 2008-03-22 09:04:52
PST ---
Created an attachment (id=21700)
--> (https://issues.apache.org/bugzilla/attachment.cgi?id=21700)
test case for Reader buffering
war for the test
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
Remy Maucherat <[EMAIL PROTECTED]> changed:
What|Removed |Added
Status|ASSIGNED|NEEDINFO
-
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #18 from Filip Hanik <[EMAIL PROTECTED]> 2008-03-21 10:24:06 PST
---
I posted to the dev lists about this, basically, one can just take advantage of
java.io.BufferedReader to do the caching
public void doFilter(Servlet
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #17 from Filip Hanik <[EMAIL PROTECTED]> 2008-03-21 08:48:11 PST
---
(In reply to comment #15)
> Remy and Filip, thank you for immediate checks.
>
> Yes, the logic is same in all of them.
>
> (In reply to comment #11, #1
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #16 from Suzuki Yuichiro <[EMAIL PROTECTED]> 2008-03-20 19:57:35
PST ---
(In reply to comment #10)
> Not bad, but I have some problem believing that no longer using bb.getLength()
> as the value for limit (or not using Int
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #15 from Suzuki Yuichiro <[EMAIL PROTECTED]> 2008-03-20 18:26:37
PST ---
Remy and Filip, thank you for immediate checks.
Yes, the logic is same in all of them.
(In reply to comment #11, #12)
> I tested the patch on 5.5 a
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #14 from Filip Hanik <[EMAIL PROTECTED]> 2008-03-19 12:25:03 PST
---
(In reply to comment #13)
> Just use the 6.0 branch for testing, no need to test all branches, the code is
> identical in all of them. The key difference
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #13 from Remy Maucherat <[EMAIL PROTECTED]> 2008-03-19 09:55:37
PST ---
Just use the 6.0 branch for testing, no need to test all branches, the code is
identical in all of them. The key difference I see is that the limit use
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #12 from Filip Hanik <[EMAIL PROTECTED]> 2008-03-19 09:07:05 PST
---
I also tested the trunk patch, and couldn't get it to work properly
--
Configure bugmail: https://issues.apache.org/bugzilla/userprefs.cgi?tab=email
--
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #11 from Filip Hanik <[EMAIL PROTECTED]> 2008-03-19 08:58:09 PST
---
I tested the patch on 5.5 and it didn't work, I will look deeper into it
--
Configure bugmail: https://issues.apache.org/bugzilla/userprefs.cgi?tab=ema
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #10 from Remy Maucherat <[EMAIL PROTECTED]> 2008-03-19 08:38:17
PST ---
Not bad, but I have some problem believing that no longer using bb.getLength()
as the value for limit (or not using IntermediateInputStream.available()
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #9 from Suzuki Yuichiro <[EMAIL PROTECTED]> 2008-03-19 05:53:43
PST ---
Although they may be not perfect possibly,
I tested these patches on some patterns that includes above-mentioned.
--
Configure bugmail: https://issu
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #8 from Suzuki Yuichiro <[EMAIL PROTECTED]> 2008-03-19 05:15:26
PST ---
Created an attachment (id=21686)
--> (https://issues.apache.org/bugzilla/attachment.cgi?id=21686)
patch for tomcat5.5
this patch may fix some bufferi
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #7 from Suzuki Yuichiro <[EMAIL PROTECTED]> 2008-03-19 05:13:54
PST ---
Created an attachment (id=21685)
--> (https://issues.apache.org/bugzilla/attachment.cgi?id=21685)
patch for tomcat5.5
this patch may fix some bufferi
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #6 from Suzuki Yuichiro <[EMAIL PROTECTED]> 2008-03-19 05:12:19
PST ---
Created an attachment (id=21684)
--> (https://issues.apache.org/bugzilla/attachment.cgi?id=21684)
patch for tomcat6.x repository
this patch may fix s
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
--- Comment #5 from Suzuki Yuichiro <[EMAIL PROTECTED]> 2008-03-19 05:10:13
PST ---
Created an attachment (id=21683)
--> (https://issues.apache.org/bugzilla/attachment.cgi?id=21683)
patch for trunk repository
this patch may fix some
https://issues.apache.org/bugzilla/show_bug.cgi?id=44494
Suzuki Yuichiro <[EMAIL PROTECTED]> changed:
What|Removed |Added
CC||[EMAIL PROTEC
60 matches
Mail list logo
